Cross-chain is not multi-chain. A true multi-chain application requires shared state and atomic execution across domains, which isolated bridges like Stargate or Synapse cannot provide. They create fragmented liquidity and disjointed user sessions.
the-appchain-thesis-cosmos-and-polkadot
Blog
Why Cross-Chain Composability Depends on Hub-Level Infrastructure
The promise of a multi-chain future is broken composability. True atomic execution across IBC requires a central coordinator for sequencing and security—a role the Cosmos Hub's Scheduler and Interchain Security are engineered to fill.
introduction
THE COMPOSABILITY GAP
The Multi-Chain Lie
Cross-chain applications fail because they treat bridges as simple asset pipes, ignoring the need for hub-level state synchronization.
Composability demands a hub. Protocols like Axelar and LayerZero attempt this by acting as messaging hubs, but they introduce new trust layers. Native composability requires a settlement layer, like Cosmos IBC, that validates state proofs.
The evidence is in the TVL. Applications with native cross-chain architecture, such as dYdX Chain, secure billions by controlling their own settlement. Bridged assets on Avalanche or Polygon remain siloed from DeFi logic on other chains.
The solution is standardization. The future belongs to interoperability hubs that provide verifiable state attestations, not just asset transfers. Without this, cross-chain DeFi is a collection of disconnected smart contracts.
thesis-statement
THE ARCHITECTURAL IMPERATIVE
The Hub is the Sequencer
Cross-chain composability fails without a hub-level sequencer to coordinate atomic execution across fragmented liquidity.
Hub-level sequencing is non-negotiable. A shared sequencer at the hub, like the one proposed for the Cosmos Hub or used by LayerZero's Executor, provides the global ordering required for atomic cross-chain bundles. Without it, applications must trust individual chain sequencers to cooperate, which introduces settlement risk and MEV extraction.
Composability requires atomicity. True cross-chain DeFi, like a swap on Arbitrum collateralized by an asset on Polygon, demands a coordinated execution guarantee. Hub sequencing creates a single point of failure for coordination, which is preferable to the probabilistic failure of asynchronous messaging bridges like Wormhole or CCTP.
The alternative is fragmentation. Relying on application-specific relayers, as seen with Axelar or Chainlink CCIP, balkanizes security and liquidity. This creates a winner-take-most market for the hub sequencer that can offer the cheapest, fastest atomicity for protocols like UniswapX or Aave.
Evidence: The IBC Model. The Cosmos Inter-Blockchain Communication protocol demonstrates that a hub-centric relay network enables secure, ordered packet delivery. Scaling this to a universal sequencer is the next logical step for cross-chain state transitions beyond simple asset transfers.
key-trends
WHY FRAGMENTATION KILLS COMPOSABILITY
The Appchain Coordination Crisis
The proliferation of appchains and L2s has fragmented liquidity and state, making seamless cross-chain interaction a critical infrastructure challenge.
01
The Problem: Fragmented Liquidity Pools
Every new chain creates its own isolated liquidity silos. Bridging assets manually is slow and capital-inefficient, forcing protocols to deploy and bootstrap on dozens of chains.\n- $100B+ in locked liquidity is fragmented across chains\n- ~15-30 minutes for optimistic rollup withdrawals\n- 2-5% typical slippage + bridge fees on cross-chain swaps
$100B+
Fragmented TVL
2-5%
Swap Tax
02
The Solution: Hub-Based Message Orchestration
A neutral hub like Cosmos IBC or Polygon AggLayer acts as a universal router, enabling secure, atomic state updates between appchains without direct pairwise connections.\n- ~3-6 second finality for cross-chain messages via IBC\n- Single trust assumption at the hub level, not per-chain\n- Enables native composability for protocols like Osmosis and dYdX Chain
~3-6s
Message Finality
1
Trust Layer
03
The Problem: Insecure Ad-Hoc Bridges
Relying on third-party bridges like Multichain or Wormhole introduces catastrophic systemic risk; each bridge is a new attack vector with over $2B exploited.\n- >50 active bridges with varying security models\n- $2B+ lost in bridge hacks since 2021\n- No universal standard for cross-chain verification
$2B+
Exploited
>50
Attack Vectors
04
The Solution: Shared Security & Light Clients
Hubs provide a shared security layer. Cosmos uses light client verification, while EigenLayer and Babylon are bringing Bitcoin/ETH security to appchains.\n- Cryptographic proofs replace trusted multisigs\n- Re-staked ETH secures new chains via EigenLayer AVS\n- Drives down security costs for chains like Celestia rollups
Cryptographic
Verification
-90%
Security Cost
05
The Problem: Broken User Experience
Users face a maze of RPCs, gas tokens, and failed transactions. A simple cross-chain swap requires multiple steps, wallets, and approvals.\n- 5+ manual steps for a typical cross-chain action\n- High failure rates on public RPC endpoints\n- No atomicity—transactions can fail mid-bridge
5+
Manual Steps
High
Failure Rate
06
The Solution: Intent-Based Abstraction
Protocols like UniswapX and CowSwap abstract the complexity. Users submit an intent ("swap X for Y"), and a solver network like Across or LayerZero finds the optimal route across chains.\n- Single transaction user experience\n- Competitive routing via solver auctions\n- MEV protection built into the settlement layer
1-Click
UX
Optimal
Route
INFRASTRUCTURE ARCHITECTURE
The Composability Spectrum: Hub vs. Ad-Hoc
Comparing the core architectural models for enabling cross-chain application logic, from isolated bridges to unified liquidity and state layers.
| Core Feature / Metric | Ad-Hoc Bridges (e.g., Stargate, LayerZero) | Hub-Based Liquidity (e.g., Chainlink CCIP, Axelar) | Sovereign Settlement (e.g., Polymer, Hyperlane) |
|---|---|---|---|
Unified Liquidity Layer | |||
Atomic Cross-Chain Composability (AXC) | |||
Native Gas Abstraction | |||
Protocol-Level Security Model | App-specific validators | Hub validator set | Hub validator set + attestation networks |
Developer Abstraction | Per-bridge SDK integration | Single hub SDK & messaging | Universal IBC-like protocol |
Time to Finality for Cross-Chain Call | 2-5 minutes | 1-3 minutes | < 1 minute (optimistic) |
Typical Fee Model | 0.1-0.5% of tx value | Fixed fee + gas premiums | Fixed fee + gas premiums |
State Synchronization | Message passing only | Message passing only | Light client state proofs |
deep-dive
THE HUB-AND-SPOKE REALITY
Anatomy of a Cross-Chain Transaction
Cross-chain composability is not a mesh of peer-to-peer connections but a hub-centric system dependent on shared infrastructure.
Cross-chain is hub-centric. Applications like UniswapX or CowSwap execute intents across chains, but their final settlement depends on canonical bridges like Arbitrum's L1 bridge or Optimism's Bedrock. These hubs establish the definitive state for asset transfers and message passing.
Composability requires shared state. A user's action on Arbitrum cannot natively trigger a contract on Base without a shared ledger of truth. Infrastructure like Chainlink CCIP or LayerZero's Endpoint network provides this verifiable state attestation, enabling conditional logic across domains.
The bridge dictates security. The security model of the weakest bridge determines the safety of the entire cross-chain stack. A dApp using Stargate for liquidity and Wormhole for messages inherits the risk profile of both, creating a composite attack surface.
Evidence: Over 60% of cross-chain TVL flows through the canonical bridges of Ethereum L2s and Cosmos IBC, not third-party solutions, proving that native, hub-level infrastructure is the primary trust vector for composability.
counter-argument
THE COMPOSABILITY IMPERATIVE
The Sovereign Chain Rebuttal (And Why It's Wrong)
Sovereign chains sacrifice the network effects that drive adoption by ignoring the infrastructure required for seamless cross-chain interaction.
Sovereignty breaks composability. Isolated chains create liquidity and user fragmentation, defeating the purpose of a multi-chain world. Protocols like Uniswap and Aave require unified liquidity pools and price oracles that cannot exist across disconnected state machines.
Hub-level infrastructure is non-negotiable. Secure cross-chain messaging via LayerZero or Wormhole and shared sequencing layers like Espresso Systems are prerequisites for atomic, trust-minimized transactions. Without them, users face custodial bridges and broken UX.
The market vote is clear. Over 60% of DeFi TVL resides on Ethereum L2s precisely because they inherit Ethereum's security while enabling fast, cheap transactions. Truly sovereign chains like Celestia rollups must still integrate this hub-level plumbing to compete.
Evidence: The bridge volume metric. Daily cross-chain volume via Across, Stargate, and Synapse exceeds $500M, proving demand for composability. Sovereign chains that ignore this demand cede users to interconnected ecosystems like Arbitrum and Optimism.
protocol-spotlight
HUB-LEVEL INFRASTRUCTURE
The Competitive Landscape: Who Else is Building This?
Cross-chain composability is a middleware war; the winning layer will be the one that provides the most secure, efficient, and programmable hub.
01
LayerZero: The Messaging Monolith
Aims to be the TCP/IP for blockchains, but its generalized message-passing model pushes complexity and risk to the application layer.\n- Key Benefit: Generalized programmability for any cross-chain logic.\n- Key Drawback: Application developers bear the security burden of configuring and verifying Oracle and Relayer networks.
$10B+
TVL Secured
40+
Chains
02
Axelar & Wormhole: The Specialized Secure Transport Layers
Focus on providing a verified, canonical state root for destination chains, acting as a trusted bridge for assets and data.\n- Key Benefit: Decentralized validator sets provide strong security guarantees for simple transfers.\n- Key Limitation: Composability is often an afterthought; complex, multi-step intents require additional application-layer coordination, creating fragmentation.
~3-5s
Finality
50+
Chains
03
The Problem: Application-Specific Silos (e.g., Chainlink CCIP, Circle CCTP)
These are purpose-built for specific use cases (oracles, stablecoins), creating secure but isolated corridors. True composability requires a shared, universal layer.\n- Key Benefit: Optimized security and liquidity for a single vertical.\n- Key Drawback: No shared state; a DeFi protocol cannot natively compose a CCTP transfer with a Chainlink price feed in a single atomic cross-chain transaction.
1
Use Case
High
Trust Assumption
04
The Solution: A Sovereign Execution Hub (e.g., Polymer, Hyperlane)
These projects are converging on the hub model, where a dedicated blockchain coordinates cross-chain state. This is the correct primitive for composability.\n- Key Benefit: Sovereign execution environment enables atomic multi-chain operations and shared security.\n- Key Benefit: Becomes the canonical liquidity layer, aggregating bridges like LayerZero and Axelar as connectivity modules, not primary settlement layers.
Atomic
Composability
Hub-Centric
Architecture
risk-analysis
CROSS-CHAIN FRAGILITY
The Bear Case: What Could Break the Hub Thesis?
The promise of a unified liquidity and state layer fails if core infrastructure assumptions are wrong.
01
The Liquidity Fragmentation Trap
Hubs like Cosmos and Polkadot assume shared security attracts TVL, but sovereign chains can still silo liquidity. Without a universal AMM or money market at the hub level, capital efficiency plummets.
- Osmosis and Astroport remain app-chain specific, not hub-wide.
- Bridging assets creates $2B+ in stranded, non-composable capital.
- The result is a hub-as-router, not a hub-as-unified-L1.
$2B+
Stranded TVL
-70%
Capital Efficiency
02
Sovereignty vs. Synchrony
Composability requires synchronous execution across chains. Hub architectures relying on asynchronous IBC or XCM create unhedgeable execution risk for cross-chain DeFi.
- A swap on Osmosis and a borrow on Mars Protocol cannot be atomic.
- This breaks complex intent-based flows championed by UniswapX and CowSwap.
- The hub becomes a coordination bottleneck, not a performance layer.
~2s
IBC Latency
0
Atomic Guarantees
03
The Validator Cartel Problem
Hub security depends on a validator set relaying all interchain messages. This creates a centralized liveness dependency and a single point of economic capture.
- ~150 validators secure Cosmos Hub; downtime halts the entire ecosystem.
- Relayer incentives are often misaligned, leading to stale price oracles and failed messages.
- Projects like LayerZero and Axelar exploit this by offering validator-set competition.
150
Active Validators
1
Liveness Failure Point
04
The Interoperability Standard War
Hub thesis assumes one protocol (IBC, XCM) wins. In reality, bridges like LayerZero, Wormhole, and Circle's CCTP create competing standards that fragment the hub's utility.
- Developers integrate 3+ bridging SDKs, negating the hub's 'universal' value prop.
- EVM-centric standards (e.g., ERC-7683 for intents) bypass hub-native tooling.
- The hub becomes just another spoke in a mesh network.
3+
SDKs per App
-40%
Hub Utility
05
Economic Model Collapse
Hub tokens (ATOM, DOT) derive value from security rental. If sovereign chains bypass shared security for cheaper alternatives, the hub's fee model implodes.
- Neutron (Cosmos) uses Replicated Security but many chains opt for Babylon or EigenLayer.
- Staking yields collapse without sufficient demand for block space.
- The token becomes a governance vehicle with zero cash flow.
<5%
Security Adoption
$0
Protocol Revenue
06
The UX Dead End
End-users don't care about hubs. Wallets like MetaMask and Rainbow abstract chains away, while intent-based architectures like UniswapX and Across route users to the best chain automatically.
- The hub is an infrastructure detail, not a user-facing product.
- Aggregation layers built on Ethereum or Solana can offer better composability without hub baggage.
- Hub-level innovation gets bypassed by application-layer abstraction.
1-Click
User Abstraction
0
Hub Visibility
future-outlook
THE ARCHITECTURAL SHIFT
The Endgame: Hubs as Intent-Based Networks
Cross-chain composability will be orchestrated by intent-based hubs, not by direct chain-to-chain bridges.
Hubs abstract chain complexity. Users express desired outcomes, and the hub's solver network handles routing, liquidity, and execution across chains like Arbitrum and Base. This mirrors the intent-based architecture of UniswapX and CowSwap.
Direct bridges fragment liquidity. A user bridging from Avalanche to Polygon via Stargate cannot interact with a dApp on Optimism without another hop. Hub-level infrastructure aggregates liquidity and routes intents through the optimal path.
Sovereign chains become spokes. Chains like Solana and zkSync Era act as execution layers, while hubs like LayerZero and Cosmos IBC become the coordination layer. This separates state consensus from cross-chain messaging.
Evidence: The 80% failure rate of complex multi-chain DeFi transactions today necessitates this abstraction. Hubs reduce this to a single user signature, shifting complexity to the solver network.
takeaways
WHY HUB-LEVEL INFRASTRUCTURE IS NON-NEGOTIABLE
TL;DR for Busy Builders
Cross-chain composability is broken without a shared settlement and state layer. Here's the technical reality.
01
The Fragmented State Problem
Applications like UniswapX and CowSwap need a single source of truth for cross-chain intents. Without it, you're building on probabilistic, asynchronous state across 50+ chains.\n- Guarantees: Hub-level infrastructure provides atomic composability, not just message passing.\n- Analogy: It's the difference between a global database and a network of sticky notes.
50+
Async Chains
0
Atomic Guarantees
02
Security is Not a Feature
Bridges like Across and LayerZero externalize security to their own validator sets, creating systemic risk. A hub (e.g., a shared sequencer network or L1) internalizes this cost, amortizing it across all applications.\n- Economic Security: A $10B+ TVL hub secures all apps, vs. each app securing its own bridge.\n- Attack Surface: Reduces vectors from O(n) to O(1) for cross-chain logic.
$10B+
TVL Security
O(1)
Attack Surface
03
The Liquidity Silos
Fragmented liquidity across chains kills capital efficiency. A hub acts as a canonical clearing layer, enabling native cross-chain AMMs and money markets without wrapped assets.\n- Capital Efficiency: Enables single-sided LPing across chains, not multi-chain replication.\n- Settlement Finality: Moves from minutes to ~500ms for cross-chain arbitrage and rebalancing.
~500ms
Settlement Time
1x
Capital Required
04
Developer UX is an Infrastructure Problem
Building cross-chain dApps today means integrating 5+ SDKs and managing chain-specific gas. A hub provides a single RPC endpoint and gas abstraction.\n- Abstraction: Write logic once, deploy to the hub, execute everywhere.\n- Cost: Reduces integration overhead by -70% and eliminates chain-specific devops.
1
RPC Endpoint
-70%
Dev Time
05
Interoperability != Composability
LayerZero and Wormhole move assets; they don't enable smart contracts on Chain A to trustlessly trigger state changes on Chain B. A hub enables synchronous composability where actions across chains are part of a single transaction.\n- Capability: Enables novel primitives like cross-chain MEV capture and shared order books.\n- Limitation: Pure messaging protocols cannot do this without a trusted third party.
Synchronous
Composability
0
Trusted Parties
06
The Economic Sinkhole of Gas
Paying gas on 5 chains for one user operation burns value and creates a terrible UX. A hub with a unified fee market and gas abstraction (like EIP-4337 on steroids) captures this value for the ecosystem.\n- Efficiency: ~50% cost reduction for users by batching and optimizing execution.\n- Value Capture: Fees are recycled into protocol security/staking, not lost to L1 miners.
~50%
Cost Reduced
1
Fee Market
ENQUIRY
Get In Touch
today.
Our experts will offer a free quote and a 30min call to discuss your project.
NDA Protected
Confidentiality24h Response
Time to ValueDirectly to Engineering Team
No Sales Fluff10+
Protocols Shipped
$20M+
TVL Overall